Multicast distribution tree allocation using machine learning

Abstract

In one embodiment, a device deploys a first machine learning model to an inference location in a network. The first machine learning model is used at the inference location to make inferences about the network. The device receives, from the inference location, an indication that the first machine learning model is exhibiting poor performance. The device identifies a corrective measure for the poor performance that minimizes resource consumption by a model training pipeline of the device. The device deploys, based on the corrective measure, a second machine learning model to the inference location. The second machine learning model is used in lieu of the first machine learning model to make the inferences about the network.

What is claimed is: 
     
         1 . A method comprising:
 obtaining, by a device in a network, data regarding multicast traffic in the network;   maintaining, by the device, a machine learning model configured to model traffic patterns in the network based on the data regarding the multicast traffic in the network;   identifying, by the device and using the machine learning model, a particular multicast traffic flow in the network as being of a particular traffic pattern; and   causing, by the device and based on the particular traffic pattern, a multicast distribution tree to be allocated in the network for the particular multicast traffic flow.   
     
     
         2 . The method as in  claim 1 , wherein the multicast distribution tree connects a plurality of provider edge (PE) routers in the network. 
     
     
         3 . The method as in  claim 1 , wherein identifying the particular multicast traffic flow in the network as being of a particular traffic pattern:
 using the machine learning model to predict that the particular traffic pattern will occur in the network, wherein the multicast distribution tree is allocated proactively for the particular multicast traffic flow.   
     
     
         4 . The method as in  claim 1 , wherein identifying the particular traffic pattern in the network comprises:
 receiving a request from a particular router in the network that comprises data regarding the particular multicast traffic flow; and   using the data regarding the particular multicast traffic flow as input to the machine learning model, to identify the particular multicast traffic flow as being of the particular traffic pattern.   
     
     
         5 . The method as in  claim 4 , wherein the data regarding the particular multicast traffic flow is indicative of an application type associated with the particular multicast traffic flow. 
     
     
         6 . The method as in  claim 1 , wherein causing the multicast distribution tree to be allocated in the network for the particular multicast traffic flow comprises:
 notifying each of a plurality of routers in the network regarding allocation of the multicast distribution tree.   
     
     
         7 . The method as in  claim 6 , wherein the particular multicast traffic flow is migrated from a default multicast distribution tree in the network to the multicast distribution tree. 
     
     
         8 . The method as in  claim 1 , further comprising:
 training the machine learning model to detect a new traffic pattern in the network.   
     
     
         9 . An apparatus, comprising:
 one or more network interfaces;   a processor coupled to the one or more network interfaces and configured to execute one or more processes; and   a memory configured to store a process that is executable by the processor, the process when executed configured to:
 obtain data regarding multicast traffic in a network; 
 maintain a machine learning model configured to model traffic patterns in the network based on the data regarding the multicast traffic in the network; 
 identify, using the machine learning model, a particular multicast traffic flow in the network as being of a particular traffic pattern; and 
 cause, based on the particular traffic pattern, a multicast distribution tree to be allocated in the network for the particular multicast traffic flow. 
   
     
     
         10 . The apparatus as in  claim 9 , wherein the multicast distribution tree connects a plurality of provider edge (PE) routers in the network. 
     
     
         11 . The apparatus as in  claim 9 , wherein the apparatus identifies the particular multicast traffic flow in the network as being of a particular traffic pattern by:
 using the machine learning model to predict that the particular traffic pattern will occur in the network, wherein the multicast distribution tree is allocated proactively for the particular multicast traffic flow.   
     
     
         12 . The apparatus as in  claim 9 , wherein the apparatus identifies the particular multicast traffic flow in the network as being of a particular traffic pattern by:
 receiving a request from a particular router in the network that comprises data regarding the particular multicast traffic flow; and   using the data regarding the particular multicast traffic flow as input to the machine learning model, to identify the particular multicast traffic flow as being of the particular traffic pattern.   
     
     
         13 . The apparatus as in  claim 12 , wherein the data regarding the particular multicast traffic flow is indicative of an application type associated with the particular multicast traffic flow. 
     
     
         14 . The apparatus as in  claim 9 , wherein the apparatus causes the multicast distribution tree to be allocated in the network for the particular multicast traffic flow by:
 notifying each of a plurality of routers in the network regarding allocation of the multicast distribution tree.   
     
     
         15 . The apparatus as in  claim 14 , wherein the particular multicast traffic flow is migrated from a default multicast distribution tree in the network to the multicast distribution tree. 
     
     
         16 . The apparatus as in  claim 9 , wherein the process when executed is further configured to:
 train the machine learning model to detect a new traffic pattern in the network.   
     
     
         17 . A tangible, non-transitory, computer-readable medium storing program instructions that cause a device in a network to execute a process comprising:
 obtaining, by the device, data regarding multicast traffic in the network;   maintaining, by the device, a machine learning model configured to model traffic patterns in the network based on the data regarding the multicast traffic in the network;   identifying, by the device and using the machine learning model, a particular multicast traffic flow in the network as being of a particular traffic pattern; and causing, by the device and based on the particular traffic pattern, a multicast distribution tree to be allocated in the network for the particular multicast traffic flow.   
     
     
         18 . The computer-readable medium as in  claim 17 , wherein identifying the particular multicast traffic flow in the network as being of a particular traffic pattern:
 using the machine learning model to predict that the particular traffic pattern will occur in the network, wherein the multicast distribution tree is allocated proactively for the particular multicast traffic flow.   
     
     
         19 . The computer-readable medium as in  claim 17 , wherein identifying the particular traffic pattern in the network comprises:
 receiving a request from a particular router in the network that comprises data regarding the particular multicast traffic flow; and   using the data regarding the particular multicast traffic flow as input to the machine learning model, to identify the particular multicast traffic flow as being of the particular traffic pattern.   
     
     
         20 . The computer-readable medium as in  claim 19 , wherein causing the multicast distribution tree to be allocated in the network for the particular multicast traffic flow comprises:
 notifying each of a plurality of routers in the network regarding allocation of the multicast distribution tree.